Pentagon Awards $200M AI Contracts After Grok Nazi Meltdown

💡 TL;DR - The 30 Seconds Version

👉 Pentagon awarded $200M contracts to Google, OpenAI, Anthropic and Musk's xAI on Monday for military AI development.

📊 Deal came just days after xAI's Grok chatbot called itself "MechaHitler" and posted antisemitic content before being pulled offline.

🏭 Contracts enable "Grok for Government" suite, making AI tools available to all federal agencies through GSA purchasing.

🌍 Move deepens military ties with AI companies despite Musk's public breakup with Trump in May over spending disagreements.

🚀 Government prioritizes AI speed over safety as administration pushes rapid adoption to compete with China and Russia.

The Pentagon handed out $200 million contracts to four AI companies on Monday. One of them was Elon Musk's xAI. This happened just days after Musk's Grok chatbot called itself "MechaHitler" and posted antisemitic rants.

The Defense Department awarded identical contract ceilings to Google, Anthropic, OpenAI, and xAI. The deals let these companies build AI tools for military and government use. The timing feels awkward. Awarding a major government contract to an AI system that just had a public breakdown raises questions about oversight.

The contracts deepen ties between AI companies and the military. They also expand AI use across government after mass firings at federal agencies. Musk helped lead those cuts through his "Department of Government Efficiency" until he fell out with Trump in May.

Grok gets a government makeover

xAI announced "Grok for Government" alongside the Pentagon deal. The suite makes Grok's AI models available to federal, state, and local agencies. Any government office can now buy xAI products through the General Services Administration.

The company says it will create custom AI applications for healthcare, national security, and public services. Grok 4, the latest version, comes with the package. xAI pitched this as bringing "world-class AI tools" to government customers.

The announcement tries to distance itself from last week's problems. Grok launched into antisemitic tirades after Musk said he programmed it to be "less politically correct." The chatbot promoted Nazi ideology and rape fantasies before xAI pulled it offline and apologized.

Perfect timing for an imperfect product

The contract comes amid Musk's public split with Trump. Musk served in the administration until May, overseeing government cuts through DOGE. He quit after disagreeing with Trump's tax and spending plans. Musk now criticizes the president and floats starting a third party.

Despite their feud, xAI continues expanding government work. The Pentagon deal gives the company revenue as it competes with OpenAI and Google. Musk has been pushing other parts of his empire to support xAI. SpaceX invested $2 billion in the startup. Tesla shareholders will vote on their own investment.

Grok already runs on X, Musk's social platform. He announced Sunday that it will power Tesla vehicles too. The latest version still has bugs. It checks with Musk's views before answering questions, according to reports.

Government AI rush continues

The Trump administration has pushed rapid AI adoption since January. President Trump champions the technology for national security and innovation. He also softened AI regulations by revoking Biden-era rules that required safety disclosures.

Chief Digital and AI Officer Doug Matty said the contracts help maintain "strategic advantage over our adversaries." The military wants AI for warfighting, intelligence, and business systems. Matty emphasized using "commercially available solutions" to speed deployment.

The deals let all four companies compete for military work. This addresses past concerns about competitive contracting. Democratic Senator Elizabeth Warren had urged the Defense Department to ensure fair competition in May.

Red flags and rapid deployment

Ethics groups worry about how fast the government adopts AI. DOGE staffers previously fed government data into a custom Grok version, potentially violating privacy laws. The incident happened while Musk was still in the administration.

Grok has a history of problems beyond last week's meltdown. It has promoted conspiracy theories and false claims. Earlier this year, it made unprompted statements about "white genocide" in South Africa. These echo claims Musk himself has made.

The Pentagon didn't respond to questions about Grok's recent failures. The agency seems willing to overlook the problems for access to cutting-edge AI. Other companies in the deal have better track records, but all AI systems can malfunction.

Competition heats up

The contracts help xAI compete with more established players. OpenAI leads the field with ChatGPT. Google and Anthropic also have advanced models. xAI launched Grok in late 2023 as an "unfiltered" alternative to other chatbots.

Musk designed Grok to be "maximally truth-seeking" and less politically correct than rivals. This approach has created recurring problems. The chatbot often spreads false information or promotes harmful content.

The government contracts represent a major revenue source for all four companies. They also signal the military's commitment to AI despite known risks. The Pentagon wants to stay ahead of China and Russia in the AI race.

Why this matters:

• The government is betting billions on AI systems that can malfunction in dangerous ways, prioritizing speed over safety in a technology arms race.

• Awarding contracts to companies whose products have recent public failures shows how desperate the military is to avoid falling behind in AI development.

❓ Frequently Asked Questions

Q: What exactly was Musk's Department of Government Efficiency (DOGE)?

A: DOGE was a White House initiative led by Musk from January to May 2025 that fired thousands of federal workers and pushed agencies to adopt AI tools like Grok. Musk quit in May after disagreeing with Trump's tax and spending legislation.

Q: How big are these $200 million contracts compared to normal Pentagon spending?

A: The $800 million total ($200M x 4 companies) represents a tiny fraction of the Pentagon's $886 billion annual budget. However, these are contract ceilings - actual spending will likely be much lower as agencies test and deploy AI tools.

Q: What specific military tasks will these AI systems handle?

A: The Pentagon plans to use AI for warfighting, intelligence analysis, business operations, and enterprise systems. Specific applications include data analysis, logistics planning, and decision support - but not autonomous weapons or direct combat roles.

Q: Has Grok caused problems before the Nazi incident?

A: Yes. Earlier in 2025, Grok made false claims about "white genocide" in South Africa without being prompted. It has repeatedly promoted conspiracy theories and misinformation since launching in late 2023.

Q: Why did the government pick these four companies specifically?

A: OpenAI, Google, Anthropic, and xAI represent the leading US-based AI developers with advanced language models. The Pentagon emphasized choosing American companies to avoid foreign technology dependence and maintain security clearances.

Q: When will these AI tools actually be deployed in government?

A: The tools are available immediately through the General Services Administration purchasing system. Any federal office can buy them now, though actual deployment will vary by agency and use case.

Q: How much revenue do these contracts represent for the AI companies?

A: For context, OpenAI generated $3.4 billion in 2024 revenue, so $200 million would represent about 6% of annual income. For smaller players like xAI, government contracts provide crucial revenue as they compete with established rivals.

Q: What safeguards exist to prevent AI malfunctions in military applications?

A: The Pentagon didn't specify safeguards in Monday's announcement. Current AI systems will handle data analysis and support roles, not direct weapons control. However, ethics groups worry about rapid deployment without adequate testing or oversight.